 |
|
|
|||||
December 15, 2025the deadline for enrolling in a marketplace plan through the Affordable Care Act for 2026came and went without an agreement on the federal subsidies that kept ACA plans more affordable for many Americans. Despite a last-ditch attempt in the House to extend ACA subsidies, with Congress adjourning for the year on December 19, its looking almost certain that Americans relying on ACA subsidies will face a steep increase in healthcare costs in 2026. As a gerontologist who studies the U.S. healthcare system, Im aware that disagreements about healthcare in America have a long history. The main bone of contention is whether providing healthcare is the responsibility of the government or of individuals or their employers. The ACA, passed in 2010 as the countrys first major piece of health legislation since the passage of Medicare and Medicaid in 1965, represents one more chapter in that long-standing debate. That debate explains why the health law has fueled so much political divisivenessincluding a standoff that spurred a record-breaking 43-day-long government shutdown, which began on October 1, 2025. In my view, regardless of how Congress resolves, or doesnt resolve, the current dispute over ACA subsidies, a durable U.S. healthcare policy will remain out of reach until lawmakers address the core question of who should shoulder the cost of healthcare. The ACAs roots In the years before the ACAs passage, some 49 million Americans15% of the populationlacked health insurance. This number had been rising in the wake of the 2008 recession. Thats because the majority of Americans ages 18 to 64 with health insurance receive their health benefits through their employer. In the 2008 downturn, people who lost their jobs basically lost their healthcare coverage. For those who believed government had a primary role in providing health insurance for its citizens, the growing number of people lacking coverage hit a crisis point that required an intervention. Those who place responsibility on individuals and employers saw the ACA as a perversion of the governments purpose. The political parties could find no common groundand this challenge continues. The major goal of the ACA was to reduce the number of uninsured Americans by about 30 million people, or to about 3% of the U.S. population. It got about halfway there: Today, about 26 million Americans, or 8%, are uninsured, though this number fluctuates based on changes in the economy and federal and state policy. Health insurance for all? The ACA implemented an array of strategies to accomplish this goal. Some were popular, such as allowing parents to keep their kids on their family insurance until age 26. Some were unpopular, such as the mandate that everyone must have insurance. But two strategies in particular had the biggest impact on the number of uninsured. One was expanding the Medicaid program to include workers whose income was below 138% of the poverty line. The other was providing subsidies to people with low and moderate incomes that could help them buy health insurance through the ACA marketplace, a state or federal health exchange through which consumers could choose health insurance plans. Medicaid expansion was controversial from the start. Originally, the ACA mandated it for all states, but the Supreme Court eventually ruled that it was up to each state, not the federal government, to decide whether to do so. As of December 2025, 40 states and the District of Columbia have implemented Medicaid expansion, insuring about 20 million Americans. Meanwhile, the marketplace subsidies, which were designed to help people who were working but could not access an employer-based health plan, were not especially contentious early on. Everyone receiving a subsidy was required to contribute to their insurance plans monthly premium. People earning $18,000 or less annually, which in 2010 was 115% of the income threshold set by the federal government as poverty level, contributed 2.1% of their plans cost, and those earning $60,240, which was 400% of the federal poverty level, contributed 10%. People making more than that were not eligible for subsidies at all. In 2021, legislation passed by the Biden administration to stave off the economic impact of the COVID-19 pandemic increased the subsidy that people could receive. The law eliminated premiums entirely for the lowest income people and reduced the cost for those earning more. And, unlike before, people making more than 400% of the federal poverty levelabout 10% of marketplace enrolleescould also get a subsidy. These pandemic-era subsidies are set to expire at the end of 2025. Cost versus coverage If the COVID-19-era subsidies expire, healthcare costs would increase substantially for most consumers, as ACA subsidies return to their original levels. So someone making $45,000 annually will now need to pay $360 a month for health insurance, increasing their payment by 74%, or $153 monthly. Whats more, these changes come on top of price hikes to insurance plans themselves, which are estimated to increase by about 18% in 2026. With these two factors combined, many ACA marketplace users could see their health insurance costs rise more than 100%. Some proponents of extending COVID-19-era subsidies contend that the rollback will result in an estimated 6 million to 7 million people leaving the ACA marketplace and that some 5 million of these Americans could become uninsured in 2026. Congressional gridlock over a healthcare bill continues. Policies in the tax and spending package signed into law by President Donald Trump in July 2025 are amplifying the challenge of keeping Americans insured. The Congressional Budget Office projects that the Medicaid cuts alone, stipulated in the package, may result in more than 7 million people becoming uninsured. Combined with other policy changes outlined in the law and the rollback of the ACA subsidies, that number could hit 16 million by 2034essentially wiping out the majority of gains in health insurance coverage that the ACA achieved since 2010. Subsidy downsides These enhanced ACA subsidies are so divisive now in part because they have dramatically driven up the federal governments healthcare bill. Between 2021 and 2024, the number of people receiving subsidies doubledresulting in many more people having health insurance, but also increasing federal ACA expenditures. In 2025, almost 22 million Americans who purchased a marketplace plan received a federal subsidy to help with the costs, up from 9.2 million in 2020a 137% increase. Those who oppose the extension counter that the subsidies cost the government too much and fund high earners who dont need government supportand that temporary emergencies, even ones as serious as a pandemic, should not result in permanent changes. Another critique is that employers are using the ACA to reduce their responsibility for employee coverage. Under the ACA, employers with more than 50 employees must provide health insurance, but for companies with fewer employers, that requirement is optional. In 2010, 92% of employers with 25 to 49 workers offered health insurance, but by 2025, that proportion had dropped to 64%, suggesting that companies of this size are allowing the ACA to cover their employees. Diverging solutions The U.S. has the most expensive healthcare system in the world by far. The projected increase in the number of uninsured people over the next 10 years could result in even higher costs, as fewer people get preventive care and delayed healthcare interventions, ultimately leading to more complex medical care Federal policy clearly shapes health insurance coverage, but state-level policies play a role too. Nationally, about 8% of people under age 65 were uninsured in 2023, yet that rate varied widelyfrom 3% in Massachusetts to 18.6% in Texas. States under Republican leadership on average have a higher percentage of uninsured people than do those under Democratic leadership, mirroring the political differences driving the national debate over who is responsible for shouldering the costs of healthcare. With dueling ideologies come dueling solutions. For those who believe that the government is responsible for the health of its citizens, expanding health insurance coverage and financing this expansion through taxes presents a clear approach. For those who say the burden should fall on individuals, reliance on the free market drives the fixon the premise that competition between health insurers and providers offers a more effective way to solve the cost challenges than a government intervention. Without finding resolution on this core issue, the U.S. will likely still be embroiled in this same debate for years, if not decades, to come. Robert Applebaum is a senior research scholar in gerontology at Miami University. This article is republished from The Conversation under a Creative Commons license. Read the original article.
Category:
E-Commerce
People and institutions are grappling with the consequences of AI-written text. Teachers want to know whether students work reflects their own understanding; consumers want to know whether an advertisement was written by a human or a machine. Writing rules to govern the use of AI-generated content is relatively easy. Enforcing them depends on something much harder: reliably detecting whether a piece of text was generated by artificial intelligence. Some studies have investigated whether humans can detect AI-generated text. For example, people who themselves use AI writing tools heavily have been shown to accurately detect AI-written text. A panel of human evaluators can even outperform automated tools in a controlled setting. However, such expertise is not widespread, and individual judgment can be inconsistent. Institutions that need consistency at a large scale therefore turn to automated AI text detectors. The problem of AI text detection The basic workflow behind AI text detection is easy to describe. Start with a piece of text whose origin you want to determine. Then apply a detection tool, often an AI system itself, that analyzes the text and produces a score, usually expressed as a probability, indicating how likely the text is to have been AI-generated. Use the score to inform downstream decisions, such as whether to impose a penalty for violating a rule. This simple description, however, hides a great deal of complexity. It glosses over a number of background assumptions that need to be made explicit. Do you know which AI tools might have plausibly been used to generate the text? What kind of access do you have to these tools? Can you run them yourself, or inspect their inner workings? How much text do you have? Do you have a single text or a collection of writings gathered over time? What AI detection tools can and cannot tell you depends critically on the answers to questions like these. There is one additional detail that is especially important: Did the AI system that generated the text deliberately embed markers to make later detection easier? These indicators are known as watermarks. Watermarked text looks like ordinary text, but the markers are embedded in subtle ways that do not reveal themselves to casual inspection. Someone with the right key can later check for the presence of these markers and verify that the text came from a watermarked AI-generated source. This approach, however, relies on cooperation from AI vendors and is not always available. How AI text detection tools work One obvious approach is to use AI itself to detect AI-written text. The idea is straightforward. Start by collecting a large corpus, meaning collection of writing, of examples labeled as human-written or AI-generated, then train a model to distinguish between the two. In effect, AI text detection is treated as a standard classification problem, similar in spirit to spam filtering. Once trained, the detector examines new text and predicts whether it more closely resembles the AI-generated examples or the human-written ones it has seen before. The learned-detector approach can work even if you know little about which AI tools might have generated the text. The main requirement is that the training corpus be diverse enough to include outputs from a wide range of AI systems. But if you do have access to the AI tools you are concerned about, a different approach becomes possible. This second strategy does not rely on collecting large labeled datasets or training a separate detector. Instead, it looks for statistical signals in the text, often in relation to how specific AI models generate language, to assess whether the text is likely to be AI-generated. For example, some methods examine the probability that an AI model assigns to a piece of text. If the model assigns an unusually high probability to the exact sequence of words, this can be a signal that the text was, in fact, generated by that model. Finally, in the case of text that is generated by an AI system that embeds a watermark, the problem shifts from detection to verification. Using a secret key provided by the AI vendor, a verification tool can assess whether the text is consistent with having been generated by a watermarked system. This approach relies on information that is not available from the text alone, rather than on inferences drawn from the text itself. AI engineer Tom Dekan demonstrates how easily commercial AI text detectors can be defeated. Limitations of detection tools Each family of tools comes with its own limitations, making it difficult to declare a clear winner. Learning-based detectors, for example, are sensitive to how closely new text resembles the data they were trained on. Their accuracy drops when the text differs substantially from the training corpus, which can quickly become outdated as new AI models are released. Continually curating fresh data and retraining detectors is costly, and detectors inevitably lag behind the systems they are meant to identify. Statistical tests face a different set of constraints. Many rely on assumptions about how specific AI models generate text, or on access to those models probability distributions. When models are proprietary, frequently updated or simply unknown, these assumptions break down. As a result, methods that work well in controlled settings can become unreliable or inapplicable in the real world. Watermarking shifts the problem from detection to verification, but it introduces its own dependencies. It relies on cooperation from AI vendors and applies only to text generated with watermarking enabled. More broadly, AI text detection is part of an escalating arms race. Detection tools must be publicly available to be useful, but that same transparency enables evasion. As AI text generators grow more capable and evasion techniques more sophisticated, detectors are unlikely to gain a lasting upper hand. Hard reality The problem of AI text detection is simple to state but hard to solve reliably. Institutions with rules governing the use of AI-written text cannot rely on detection tools alone for enforcement. As society adapts to generative AI, we are likely to refine norms around acceptable use of AI-generated text and improve detection techniques. But ultimately, well have to learn to live with the fact that such tools will never be perfect. Ambuj Tewari is a professor of statistics a the University of Michigan. This article is republished from The Conversation under a Creative Commons license. Read the original article.
Category:
E-Commerce
Below, Nicholas Thompson shares five key insights from his new book, The Running Ground: A Father, a Son, and the Simplest of Sports. Thompson is CEO of The Atlantic. In his time as CEO, the company has seen record subscriber growth. Before this role, he was editor-in-chief of Wired magazine. He is also a former contributor for CBS News and has previously served as editor. As a runner, he set the American record for men ages 45-plus in the 50K race. Whats the big idea? Running has the capacity to show us what were made of and help us grow beyond our limitsboth as we race ahead on the track and in life. Struggle, aging, and even trauma can become engines of transformation if we learn how best to keep moving forward. 1. You dont stop running because you get old. You get old because you stop running. I used to think that you would just get better and better with age until youre about 28, and then you would get worse and worse. But as Ive gotten older, Ive learned that isnt true. In fact, I ran my fastest marathon at age 44. Of course, there are certain things that decline in a runners life, as they do for everyone. Over the years, your bone density deteriorates, your VO2 max goes down, and youre more likely to get a little injured here or there. But while that happens, there are things that get better. We gain mitochondrial efficiency, for example, and most importantly, we get wiser. We have learned more about training. We have learned more about our limits. And not only that, but we can also pick up new habits to do things differently. In some ways, aging is like youre on a moving sidewalk that is going backwards, but youre picking up things that allow you to go forward. If you can go the same speed forwards as youre going backwards, then you run the same time year after yearwhich is what I did in my thirties. But sometimes, you can actually get better by going forward faster on that sidewalk than its pushing you backwardand thats what I did in my mid-forties. This applies beyond running. I had this conversation with my mother recently: Shes in her mid-seventies, and she said, Nick, my reflexes are just getting worse and worse with age. I said, There are things that are going to make your reflexes worse or worse with age, but what if we tried to go the other direction? Then I got her out on our front porch and I started tossing her tennis balls, and she started catching them. I tossed them a little more to the side, and it turned out that her reflexes could get better. Yes, aging is real. Unquestionably. There are many forces that slow us down, but what slows us down the most is when we give in and say, I dont want to do it today. When that happens, thats when you really start to slow down. Thats when you start to age. What you should do is push back as best you can. 2. Most pain is just a prediction. When I was a young runner, I believed that pain was purely physiological. I would exercise, my body would produce lactic acid, and the lactic acid would somehow trigger fatigue or your muscles would micro-tear and that would trigger pain signals. But as I got older, I read more studies, thought more, read the work of people like Alex Hutchinson and Tim Noakes, and realized that pain is something quite different when you run a race. Pain is weird. It moves all over the body. Maybe Ill feel it in my calf and then my quad, and then Ill feel like I have an upset stomach or Im nauseous, or dizzy, or experience general malaise. Maybe my shoulder will hurt. Whats going on? Its not that theres actually something wrong in my quad and then my knees and then my stomach. This is my brain having a conversation with the rest of my body. The brain is worried about losing homeostasis. It doesnt think that I can run this speed for this long. Maybe it doesnt think I can run 26.2 miles on this hot day, at this particular speed, and so its trying to slow me down because it doesnt want to enter a state where it could be at risk. During a race, pain is the brain trying to convince the body to slow down. If thats true, what does that mean about training? First, you should try to reset your brains expectations so that it doesnt get so scared. When Im in a marathon training cycle, I know that I cant run every day as hard as Im going to run on race day. But I try to stress each system in the body more on one day during the training cycle than I plan to on race day. Maybe that means using a single training day to run more than 26.2 miles. Maybe I run 20 miles while dehydrated. Maybe I will run 15 miles down a mountain to put extra stress on the quads. Its a way of getting the brain to understand that those levels of pain do not put me at risk. There are other things you can do, too. What resets the brains expectations when its hot? I like to rub ice on my wrist. This makes me feel a little cooler and a little better, but its also a way of resetting my brains expectations of what the temperature risks are. The great runner Eliud Kipchoge smiles when he starts to hurt. Its his way of trying to trick himself into feeling like hes okay and not worrying so much, and then the pain in the rest of his body can disappear. When running a 100K recently, I banged my toe against a root. My toenail split and stuck upthat hurt. That was real pain. That was physiological pain born of shouting nerve signals. I started to run, and I got really worried that maybe I couldnt travel the remaining distance. I think it was seven miles maybe, and I told myself I just couldnt do that. Thats when I started to hurt all over my whole body. Everything felt wrong. But then I got to an aid station, took off my shoe, took off my sock, taped down the bloody toenail, and I realized that my toenail would be fine. Once Id realized this, my whole body felt better. I didnt have to worry that something was going to go horribly wrong. This is a good lesson for life. Its a good reminder that, lots of times, what slows us down is in our own heads. Sometimes you must set an uncomfortable pace. Sometimes you must stress yourself. Whatever it is that you want to be really successful at, you have to go harder than you think you can. You have to use one part of your brain to trick another part of your brain. I call it playing hide and seek with your mind. 3. We all contain hidden versions of ourselves. I started running in high school and joined the indoor track team winter of my sophomore year. Went out and raced the 2-mile a bunch of times, ran 11 minutes and 45 seconds, then 11 minutes and 40 seconds, and at the end of the year, I was still locked in at that pace. At that point, I thought the best I could do would be 11 minutes and 30 seconds for two miles, 5 minutes and 45 seconds each. I knew the splits around the blue track at my high school, but the final race was the New England Championships, and it was hosted at a different school. The track there was a bit different, so when the race began, I didnt know exactly how fast I was running. I couldnt make sense of the splits. When I went through a mile, somebody called out 5:25. I thought they were joking, or something was wrong. I didnt believe I could run 5:25 for a mile . . . but then I finished the race and had run 10:48. Id taken my time down by 45 seconds. I was able to run what I thought was an unrealistic goal for myself because of the fact I didnt know how fast I was running. If I had known, I wouldnt have been able to go that fast./p> The same process happened 25 years later. When I was 30, in 2005, I ran a marathon at 2 hours and 43 minutes. Shortly thereafter, I was diagnosed with thyroid cancer. I went through a terrifying treatment period. I knew I would survive. It wasnt the worst kind of cancer, but it was still scary, especially at 30 years old. Afterwards, I felt like I needed to run another marathon. So, two years later I ran the New York City Marathon again in exactly 2 hours and 43 minutes. For the next 11 years, I continued to run marathons at almost exactly 2 hours and 43 minutes. In fact, I had the nickname Mr. 2-4-3. But then in my mid-forties, I started training differently. I had a coach who had me train faster, do shorter workouts, do sprints, eat a little differently, and I ended up running at 2 hours and 29 minutes. This was a completely different level of success. Why was I able to run these marathons in 2 hours and 29 minutes in my mid mid-forties, but my personal best was 2 hours and 43 minutes in my late twenties? One day, I was running across the Brooklyn Bridge and realized that I hadnt gone faster than 2 hours and 43 minutes in my thirties because thats not what I had wanted. All I had wanted to do was to go as fast as I had run before I got sick. I needed someone to reframe my expectations, to tell me that there was a faster Nick inside of me to help me. That push from my coach helped me understand that I could actually be more than I had been before I was sick. This got me to believe in myself at some deep level, and then I could run it. Sometimes our limits are in our heads. We only think we can go so far. We truly believe that limit, but we have to unlock it to go further. Maybe we can unlock it ourselves. Maybe somebody else unlocks it. Theres a different version locked inside of you who can be found. 4. You can reach transcendence through restraints. Ive always wanted to reach a level of transcendenceto step outside of the body I live in during the day, to break outside of the Nick whose mind is wired to his desk and to-do list focused. I wanted to feel like Ive reached a new spiritual plane and a deeper understanding of the world. To feel more at one with the universe, I run up mountains: as the sun comes up, deep in the forest, even losing track of where I am. Its a beautiful, glorious experience. But as I worked on the book, I realized that there are runners who are reach transcendence in almost the opposite way. I spent a lot of time with an amazing runner named Suprabha Beckjord, who won the 3,100-mile race in Queens, New York, for nine consecutive years. The way that race works is you run around a single block all day, every day. We run clockwise one day, counterclockwise the other. You start at six in the morning with a minute of meditation, and then have to be done by midnight. You go home and sleep until start time the next morning. You return to the track and do it over and over again. The race starts in August and ends in October. One person said, Its not a real race unless you have to get your hair cut in it. One year, somebody had their visa expire in the middle of the race. Suprabha taught me an important lesson. When running around the same block over and over, if you start thinking about your surroundings and what youre doing, youll go crazy. So, you learn mental practices. You learn to imagine that you are a child running in the woods. You learn to escape the boundaries of where you are. You learn to think at a much deeper level. You learn to meditate as you run. I also spent time writing about a runner named Michael Westphal. He lived on Great Cranberry Island, Maine, which has a population of about 40 people. Of that tiny population, six of their people became sub-three-hour marathoners. They ran on the same beautiful two-mile road, back and forth, back and forth, back and forth. Because of the tiny island, because of the tiny community, because of the restraints on what could be done there, they were able to reach a level of excellence. Westphal also taught me by a different kind of constraint. Later in his life, he was diagnosed with Parkinsons disease. First, he wanted to hide it, found it embarrassing, but then he realized that he loved to run, and he was going to run despite having Parkinsons. He figured out a way to run with his illness by tying his hands behind his back with string. He learned a whole new way of running. It was a different kind of restraint. It taught him humility and a sense of connection with other runners. He said something beautiful to me: Theres more to running than just beating people. You can reach transcendence through restraints. 5. Post-traumatic growth can be a subtle but serious competitive advantage. Not long ago, I was with Arthur Brooks. He writes about happiness for The Atlantic. Hes a real scholar of the field, and I asked him, Arthur, whats the number one thing that can make someone happy and content in life? He said, Well, its a weird one and you cant really force it. I said, Okay, what is it? And he said, Get cancer and survive it. When he said that, a light bulb went off. In my twenties, my running and my work was kind of a mess. As a runner, I was trying to break three hours in the marathon. That had been my fathers goal. He had come close, but not achieved it. I didnt come close at all. I ran marathon after marathon, sometimes dropping out or walking the second half. As for work, I got fired from my first job in less than an hour. My second job, I was almost fired before I started. I struggled and struggled, then had one brief period of success at a place called The Washington Monthly. But after that, I couldnt get freelance gigs. I applied for hundreds of jobs in my late twenties. I was making more money as a street musician playing guitar on the L Train than I was as a journalist. In my thirties and forties, everything got better on both those fronts. I ran much faster. My work got much better: great job at Wired, great job with The New Yorker, wonderful job right now helping run The Atlantic. In between, there was when I got thyroid cancer and faced death for the first time in my life. What Arthur Brooks said and what the studies show is that if you stand at the precipice of death and walk away, you take life more seriously afterward. To me, I think what happened was somewhat paradoxical. After my cancer experience, my goals narrowed in some ways. Instead of constantly shooting for the moon and thinking I should have everything all at once, I became more methodical about just doing what I could every day. This is the trick to running successfully, too. Yes, you do absolutely have to push yourself if you want to get better, but the most important part is learning to run every day. No matter what the weather is, no matter how you feel, no matter how much time you haveyou just go out and do it. I took that attitude toward running and work. I began asking myself, What is the best thing that I can do today? How can I do my job better today than I did it yesterday? That attitude change came partly from my thyroid cancer journey, but there are different ways people can go through an experience like that. Not just cancer, but medical scares or personal scares. When you come out the other side, you can make choices that lead to more success in whatever you set your mind to. Enjoy our full library of Book Bitesread by the authors!in the Next Big Idea app. This article originally appeared in Next Big Idea Club magazine and is reprinted with permission.
Category:
E-Commerce
All news |
||||||||||||||||||
|
||||||||||||||||||